Section One: K-5 ENROLLMENTS. Actual and Projected K-5 Enrollments

Size: px

Start display at page:

Download "Section One: K-5 ENROLLMENTS. Actual and Projected K-5 Enrollments"

Ada Dorsey
5 years ago
Views:

1 New K-5 School Enrollment Projections Still Appear Optimistic BOE Urgency to Add More K-5 Capacity Makes No Sense Connecticut s Growing Fiscal Crisis Should Restrain All Local Funding Decisions New K-5 enrollment projections are lower, but probably still too optimistic. The BOE continues to understate our K-5 school capacity by calling a 576 school a 504, and then insisting that we can put only kids in it that s basically 75% of actual capacity. Fairfield has more than enough K-5 capacity for the next several years, and perhaps much longer. There is no rush to do anything until we see what actually happens with enrollments. Connecticut faces a serious and deepening financial crisis that is likely to have significant adverse effects on our Town, including lower, if any, reimbursements for school construction. Section One: K-5 ENROLLMENTS Predicting the future is not easy, and elementary school enrollments are certainly no exception. The outcome depends on many unpredictable influences (like births and the number of families moving in and out), all of which are subject to change without notice. K-5 enrollments of 4,233 at the beginning of the current year were down 13% from 4,858 in The BOE S consultant, Milone & MacBroom (M&M), now thinks they will decline another 2% to 4,163 by , and then rise gradually to 4,378 by , still 10% below their 2009 level. These projections are important because we will be making decisions based on them that can cost taxpayers tens of millions of dollars. So, it seems reassuring when M&M says its projections are supported by demographic, housing and economic data, and that they are based on a cohort survival methodology that is accepted by CSDE School Construction Projects (CGS ), and that they include persistency ratios [that] account for the various external factors affecting enrollments, including housing characteristics, residential development, economic conditions, student transfers in and out of the system, and student mobility. However, these claims seem far less reassuring when one learns 5,500 that the same consultant, less 5,300 5,100 than two years ago, using the 4,900 same methodology, produced 4,700 projections that were quite 4,500 different. As is apparent in the 4,300 graph on the right, M&M s latest 4, projection (blue line) is 3,900 3,700 8% below (350 fewer students) 3,500 what they previously thought would happen (orange). Also shown in the graph are the even more optimistic projections (red) of another BOE consultant, MGT, that were based on equally sophisticated methodologies. Actual and Projected K-5 Enrollments Actual Old M&M New M&M MGT

2 2 In reality, the consultants assemble lots of historical data and then they guess. In the case of M&M, they guess: (a) what births will be in the future; (b) what the relationship between births and kindergarten enrollments will be (the so-called Birth-to-K ratio ); and finally, (c) how many additional students will migrate into our schools for grades 1-5 (the so-called Persistency ratio ). Implicit in M&M s methodology is a presumption that there are stable and reliable relationships between births and kindergarten enrollment, between past and future birth-to-k ratios, and between past and future persistency ratios. In fact, the historical record makes it very clear that consultants can t predict births, and that the statistical ratios upon which they rely fluctuate widely for reasons they cannot explain. Births: M&M s latest best guess about births is that, having dropped 17% from an average of 724 in the period to 598 on average in the period, and having dropped another 14% to 513 in the period, Fairfield births will rebound 13% to average 578 in the period. M&M offers no rationale for why births have declined or why they should increase. Birth-to-K Ratio: M&M s latest guess is that the BTK ratio, after averaging (within a wide range of to 1.331) 1 over the past 13 years (FY2006-FY2018), will average over the next ten years (FY19-FY28). At 1.245, M&M s guess is at the 77 th percentile of the historical range. M&M offers no rationale for why the BTK ratio has increased, or why it should remain much higher. Persistency Ratio: M&M s latest guess is that the overall Persistency (In-Migration) Ratio, after averaging (within a range of to 1.068) over the past 12 years (FY2007-FY2018), will average over the next ten years (FY2019-FY2028). At 1.068, M&M s guess is at the 100 th percentile of the historical range. M&M offers no rationale for why the Persistency ratio has increased, or why it should remain at the high end of its historical range. 2 M&M provides three different scenarios for total school enrollment, which they label as Maximum, Best Fit and Minimum. The differences are based primarily in different assumptions about future births, which they claim to have developed from regression models based on assumed unemployment rate and single-family home sales. 3 They do not provide a breakdown of their three projections into K- 5, Middle School and High School, but using their data, we can look at different alternative scenarios. One obvious scenario to consider is: What happens if births do not increase from their latest 5-year average, and what if the BTK and Persistency ratios revert to their historical means instead of staying at or near their historical peaks? The answer is that K-5 enrollments in would be 16% (713 students) lower (3,665) than M&M is now projecting (4,378), and 25% below 2009 s level (4,858). Two things seem very clear: (1) M&M s projections should be viewed with considerable skepticism; and (2) M&M s Best Fit projections are probably high because they assume substantial and sustained increases in BTK and Persistency ratios. M&M itself concedes that their methodology works well for stable populations, including communities that are growing or declining at a steady rate, but they fail to acknowledge that these conditions do not currently apply in Fairfield (or perhaps anywhere else in CT). 1 Note that this range means that at the low end, K enrollments were 4.8% below births five years earlier, and at the high end, K enrollments were 33.1% higher than births five years earlier. Accordingly, this does not appear to be a useful predictive tool. 2 M&M also does not explain why the Persistency Ratio should increase only in the first three grades (K-1, 1-2 and 2-3) and not in the final two grades (3-4 and 4-5). 3 M&M says their assumed increases in births are +23%, +10% and -5% for their maximum, best-fit and minimum projections, respectively, although it is not clear how they computed these percentage increases.

3 Meanwhile, although the M&M projections seem high, they are consistent with and therefore reinforce the fervent beliefs of several past and present BOE members that enrollments have declined in the

3 3 Meanwhile, although the M&M projections seem high, they are consistent with and therefore reinforce the fervent beliefs of several past and present BOE members that enrollments have declined in the past, but they always recover. In her Proposed Budget for , Supt. Jones provides the graph on the right with 58 years of enrollment data, which shows that elementary school enrollments did decline from a 1970 peak over 6,000 to a 1984 trough of less than 3,000, and then partially recovered to almost 5,000 in Whether they think the forces that produced this outcome are still relevant, or that there are new forces at work that will produce the same outcome has not been shared by those who assert that enrollments always recover. The Unanswered Questions What we would really like to know is why births have declined in recent years, and why the number of students who have migrated into our schools seems to have increased relative to the number of births. At this point, we can only suggest some plausible hypotheses. Births Assuming the historical data are correct, the number of Fairfield births between 2000 and 2006 averaged 725 (within a range of 638 to 841). Births then declined 18% to an average of 598 (in a tight range) in the TOTAL K-5 ENROLLMENTS School Old New M&M Alternate Scenario Year Actual M&M M&M Change Scenario* v. M&M , , , , , , , , , , ,462 4, ,421 4, ,233 4, ,018 4, , ,967 4, , ,993 4, , ,076 4, , ,205 4, , ,383 4, , ,532 4, , ,613 4, , ,324 na 3, ,378 na 3, * This scenario simply assumes that births for the next five years remain at their the last five years, and that the BTK and Persistency ratios revert to their histori instead of increasing substantially as M&M has assumed.

4 period and declined another 14% (a cumulative 29%) to an average of 513 over the next eight years (in another relatively tight range). Perhaps the most likely explanation is that the Great Recession, which began in late 2008, played an important role by discouraging childbearing. Perhaps the sustained weakness in the housing market continued the trend to smaller families. In any case, there is no obvious reason (e.g., an expected increase in population) to believe that circumstances have changed and that births will average 578 over the next five years (13% higher than the last eight years) as M&M s best-guess projections assume they will. If anything, Connecticut s deepening financial crisis and the new federal tax laws restricting the deductibility of state and local taxes, which will also affect the deductibility of mortgage interest, seem to argue for a continued decline in births. According to one recent article, The share of women who have children could drop again if current trends continue. Women are planning to have children at later ages, when they are more likely to have trouble conceiving. And the fertility rate has not rebounded after the recession in the way that many economists expected: The number of babies born per 1,000 women of childbearing age in 2016, the last year for which we have official data, was a record low. 4 The BTK Ratio As is apparent in the table on the right, the number of kids enrolled in kindergarten for the academic years averaged about 3% (1.032) more than the number of births five years earlier. This Birth-to-Kindergarten ratio then bumped up to about 15% (1.151) for the next six academic years ( ), and then bumped up again to 29% (1.288) in the last three years. A small portion of the recent bump reflects the fact that we began to admit Open Choice students into Kindergarten classes beginning in (5, 7, 2, 6 and 1 students, respectively). Perhaps, once again, the Great Recession is responsible, by forcing many more Fairfield families to choose public schools rather than private schools. Perhaps the latest bump reflects a one-time increase in the number of families with schoolage children because many seniors who had been trying to sell their homes were finally able to do Birth-to-Kindergarten Ratios Birth School K BTK Year Births Year Enrllmnt Ratio so. Perhaps the families moving into what are now heavily discounted (and thus more affordable) higher-end homes are less inclined to send their children to private schools than the prior owners. Perhaps the Millennials are finally moving to suburbs and starting families, albeit smaller families than previous generations. Perhaps the income demographics of Fairfield are changing permanently in the direction of more lower-end families and fewer higher-end families. Perhaps there has been a shift in 4 ion=topbar &WT.nav=searchWidget&module=SearchSubmit&pgtype=Homepage

5 5 the girl-boy mix of births in favor of more girls, which would temporarily increase the BTK ratio because boys are often held back a year. The Persistency Ratio Meanwhile, the Persistency (or In-Migration ) Ratio, which measures how many more students are enrolled in our elementary schools than were enrolled in Kindergarten for the previous five years, was quite stable for ten years ( ) at around 4% (1.038). However, in the last two years it too has bumped up to almost 7% (1.066). Once again, a small portion of the recent bump can be explained by changes in Open Choice admissions. For example, in , we admitted 15 O.C. students into grade one, for a total of 17, including the two students admitted to Kindergarten the previous year. This alone kicked the K-1 persistency ratio that year from to 1.047, but this does not explain why the K-1 ratio stayed at the following year when there was a net reduction of one O.C. student (from 6 to 5) in grade 1 versus Kindergarten. As noted above on the subject of births, Connecticut s deepening financial crisis, weak economic growth and the new federal tax laws restricting the deductibility of state and local taxes and of mortgage interest, would all seem to argue for reduced in-migration, but we are not aware of any valid means of predicting what will actually happen. Whatever the outcome is for K-5 enrollments, the good news for taxpayers (as we will discuss in Section Two), is that we have more than enough capacity in our K-5 schools to allow us to wait and see what actually happens to enrollments before we have to make any decisions about whether we should expand or shrink our K-5 school system. Section Two: K-5 SCHOOL CAPACITY You might think that computing capacity utilization is easy: 1. Count the number of classrooms; 2. Multiply the number of classrooms by the number of students each of them will hold to determine how many seats we have; and 3. Divide the number of students we have by the number of seats available. Unfortunately, it seems nothing is easy when it comes to understanding our $200 million school system. Counting Classrooms Educational Specifications PERSISTENCY (IN-MIGRATION) RATIO DATA School Prev. 5 Yrs. Grades 1-5 Persistency Year K Enroll. Enroll. Ratio Educational Specifications ( ed-specs ) describe what rooms, room sizes, room uses and design characteristics are deemed to be necessary by the BOE to achieve its educational goals and objectives. Readers will probably not be surprised to learn that the ed-specs have steadily increased the amount of

6 6 space that the BOE says is required in our K-5 schools. In addition to classrooms, the BOE now specifies a need for Resource (Special Education) rooms, Science rooms, Art rooms, Music rooms, Music Instrumental rooms, Gifted rooms, Occupational and Physical Therapy rooms, Conference rooms, Office rooms, and Small Groups rooms, Media Center rooms, Computer Lab rooms, Administrative rooms, and Nurses rooms, in various sizes and combinations. The latest detailed ed specs from the BOE were those created for the planned $18-$19 million renovation and expansion of Holland Hill School, a summary of which is provided in the slide on the right from a presentation by M&M dated October 24, We will not review any of these ed specs in detail, but they are important to understanding why, as you can see in the table below, there are more classrooms in each of our elementary schools than are included in the number of classrooms that are useable for instruction. CAPACITY UTILIZATION IN K-5 SCHOOLS Instructional Total Current Empty Capacity Classrooms Rooms Seats (iii) Pupils Seats Utilization Total CLC Pre-K M&M Actual M&M 95% 95% M&M 95% Pupils/Room Burr % 69% 72% Dwight % 77% 81% Holland (i) % 66% 69% Jennings % 71% 75% McKinley % 76% 80% Mill Hill (ii) % 81% 85% N. Stratfield % 66% 70% Osborn % 81% 86% Riverfield % 72% 75% Sherman % 89% 94% Stratfield % 71% 75% % 74% 78% Source: M&M at 10/24/17. (i) After current renovations and expansion. (ii) M&M numbers exclude five portables. (iii) The M&M source data include CLC seats at 8/room and exclude Pre-K seats, which makes no sense. We have excluded both CLC and Pre-K (a non-mandated program) from seats available in Instructional classrooms from all columns. 5

7 7 Counting Classrooms -- Portables The next step in determining capacity is to decide whether to include portable classrooms in the totals. In its most recent analysis, the BOE arbitrarily chose to exclude the five portables that are currently in use at Mill Hill School. We have included them because they exist and are currently in use or available for use, and because the BOE itself has always included them in the past, including last year s budget book (page 163), which shows Mill Hill s seat capacity at 378 (including the five portables). This year s book (page 167) shows capacity at only 273 ( = 105 / 21 = 5 portables). Counting Classrooms Pre-Kindergarten Space The BOE also excludes three classrooms (one at Burr and two at Stratfield), two of which are currently used for a non-mandated Pre-Kindergarten (PK) program (for only 72 students) that represents an ineffectual response to the longstanding racial imbalance problem at McKinley School. These PK classrooms will not be necessary when the racial imbalance problem is solved, as it inevitably will be, by redistricting, and thus we have included them in our total. 6 The second PK classroom at Stratfield is used for an Early Childhood Center program that could be moved elsewhere. We have not included in our instructional classroom count any other rooms that are excluded by the BOE, such as the five classrooms that are dedicated to the districtwide CLC (Complex Learner Cohort) program. Seats Per Classroom BOE Guidelines Perhaps the most important influence on any analysis of capacity utilization is how many pupils we should assume can be accommodated in each classroom. The maximum number of students is determined by the BOE s own guidelines on class sizes. These guidelines are found on page 134 the Superintendent s Budget Proposal. 7 The guidelines are: 23 for grades K-2 and 25 for grades 3-5 in all schools except McKinley, where they are in both cases, two students less. The number of students can exceed these guidelines, but the guidelines go on to specify that a class not exceeding 25 shall be desirable and that when a K-2 class exceeds 30 or a 3-6 class exceeds 35, the class many be divided or a teaching aide and/or intern provided. In the table above, we have assumed that 100% capacity is 23 students in a K-2 class and 25 students in a 3-5 class, representing an average of 24 students per K-5 classroom, though if it were necessary to do so, the guidelines would allow for more. Seats Per Classroom Loading Factor As the consulting firm, MGT, explains in its December 2010 report: Elementary schools typically group students by grade level where each class contains students of one grade. Realistically, students do not come in even groups for each grade. Consequently, it is unrealistic to expect each classroom to be filled with the maximum number of students allowed in the loading factor, e.g. 25 students in every 3 rd grade room. Therefore, to arrive at a practical capacity calculation, a 95 percent scheduling/grouping factor is used to arrive at the functional capacity. 8 6 Although the State BOE has discussed mandating universal PK in Connecticut schools, doing so is extremely unlikely in the midst of a deepening fiscal crisis for the State and many of its municipalities. If PK were mandated, all towns would be given several years to figure out how to comply %20Enrollment%20Projections%20%20Elementary%20Capacity%20Study.pdf

8 8 In short, schools can t operate at 100% of actual capacity, so some cushion must be allowed to accommodate differences from year-to-year in the number of students in each grade. MGT says a 5% cushion is adequate. In contrast, Fairfield s BOE uses a 12.5% cushion, by assuming that each classroom can accommodate only 21 students, so if there are 24 instructional classrooms, they call the school a 504 (21 x 24 = 504), when it is really a 576 (24 x 24 = 576). Then, the BOE majority asserts that we must apply another 12.5% cushion on top of that by operating at only 85%-90% of their understated capacity, which means they want only 18 students on average in classrooms that can actually accommodate 24 students which is 75% of actual capacity. Section Three: Conclusions 1. With its excessive 12.5% cushion, by ignoring the five portables at Mill Hill and by assuming that we will always need three classrooms for PK, the BOE says we have only 4,851 seats available for K-5 instruction; Fairfield Taxpayer believes that, with a MGT-recommended cushion of 5%, we have at least 12% more seats (5,449). 2. With 4,252 K-5 students at present, the BOE says our current capacity utilization is 88%; Fairfield Taxpayer believes we are actually ten points lower at 78%. 3. If the M&M projection (4,163) for is correct, the BOE says our capacity utilization in five years will decline to 86%; Fairfield Taxpayer believes the actual number would be 76%. 4. If the M&M projection (4,378) for is correct, the BOE says our capacity utilization will rise to 90%; Fairfield Taxpayer believes the actual number would be 80%. 5. If Fairfield Taxpayer s Alternative Scenario for enrollments proves correct, we will have only 3,665 students in , and a capacity utilization of only 76% based on BOE capacity, and only 67% based on 95% of actual capacity, in which case we are likely to close one or more schools. 6. We have more than enough capacity under any reasonable scenario, and more than enough time, to wait and see what actually happens to K-5 enrollments before we make any decisions about renovating and/or expanding Mill Hill School at a cost that could exceed $25 million. K-5 Enrollments vs. Capacity Actual M&M Possible BOE Capacity 95% Actual Capacity

9 BIRTH, BIRTH-TO-K RATIOS, AND KINDERGARTEN ENROLLMENT DATA PERSISTENCY (IN-MIGRATION) RATIO DATA Births Birth-to-Kindergarten Ratios K Enrollments Prev. 5 Yrs K Enrollment In-Migration Ratio 1-5 Enrollments Birth SDPH* Old New M&M School Old New M&M 1-Oct Old New M&M Old New Old New 1-Oct Old New M&M Year Actual M&M M&M Change Year Actual M&M M&M Change Actual M&M M&M Change Actual M&M M&M Actual M&M M&M Actual M&M M&M Change , , , , , , , , , , , , , , , , , , , ,660 3, ,817 3, ,535 3, ,775 3, ,381 3,352 3, ,602 3, % ,244 3, ,407 3, % ,133 3, ,281 3, % ,134 3, ,289 3, % ,189 3, ,346 3, % ,306 3, ,468 3, % ,468 3, ,638 3, % ,602 3, ,778 3, % ,670 3, ,851 3, na na 3,728 3, ,608 na na na 3, ,649 na Avg # 614.2# 578.4# * State Dept. of Public Health # Latest 5-year averages. Old M&M data are as of 2/11/16. New M&M data are as of 11/28/17.

10 10 ALTERNATE SCENARIO WITH NO CHANGE IN BIRTHS AND WITH BTK AND PERSISTENCY RATIOS AT THEIR HISTORICAL AVERAGES Births BTK Ratios K Enrollments Last 5 Yrs. K Enrollment Persistency Ratio 1-5 Enrollments K-5 Enrollments Birth SDPH* Alternate School Alternate Alternate Alternate Alternate Alternate Alternate Year Actual Scenario Year Actual Scenario Actual Scenario Actual Scenario Actual Scenario Actual Scenario Actual Scenario Avg # 521# * State Dept. of Public Health # Latest 5-year averages.

11 School PERSISTENCY RATIOS BY GRADE (Class Size Change One Year to Next) Year K E E E E E E E E E E Averages 2003-' ' '28E Assumed Increase 2.8% to 3.5% 0.3% to 1.9% 1.0% to 2.4% 0.4% to 0.1% 0.3% to 0.6% Source: M&M as of 11/29/2017, with ratios for AYs 17, 18 and 19 updated for actual enrollments. February 7, 2018

Budget. Enrollment Debt Service March 13, 2017

Budget. Enrollment Debt Service March 13, 2017 2017-18 Budget Enrollment Debt Service March 13, 2017 1 Mission-Focused Budgeting All students are provided the opportunity and necessary support to engage in relevant, challenging work which contributes